July 14, 2025 • 13 mins
In this episode of Hashtag Trending, host Jim Love discusses Intel's dramatic fall from the top 10 semiconductor companies as revealed by CEO Lip-Bu Tan. He also delves into the growing conflicts between data centers and local communities, highlighting environmental and health issues. OpenAI's failed $3 billion acquisition of Windsurf and the intriguing 'reverse acquisition' by Google is another key topic, shedding light on the rising tension in the AI industry. Finally, Love covers the astonishing performance of an AI in solving advanced mathematical problems, as reported by Scientific American, and provides an update on Rogers' Starlink satellite service test in Canada.
00:00 Introduction and Headlines
00:27 Intel's Struggles and Future Plans
02:58 Data Centers: The Hidden Costs
06:27 OpenAI's Acquisition Drama
09:13 AI Outsmarts Top Mathematicians
11:34 Rogers Starlink Satellite Service
12:39 Conclusion and Personal Note
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:04):
Intel's, CEO delivers bad news.
Data centres are coming inconflict with their neighbors.
OpenAI's, $3 billion acquisition ofWindsurf Falls apart at the last minute.
And AI shocks some of the world's bestmathematicians in a confidential review.
Welcome to Hashtag Trending.
(00:24):
I'm your host, Jim Love.
Let's get into it.
Intel's, CEO made a stunning admissionin a speech to his employees.
Intel has fallen off the list of the top10 semiconductor companies, and it's too
late for Intel to catch up with Nvidia.
Anyone listening to or readingthe tech news would be aware that
(00:46):
Intel was no longer the leader, butjust how far they've fallen was a
shock . Lip-Bu Tan the CEO made thecomments during an internal meeting
this week and they got leaked.
He said 20, 30 years ago.
We were really the leader.
Now, I think the world has changed.
We're not in the top 10semiconductor companies.
(01:11):
Now that's stunning from a companythat used to own the chip industry.
Do you remember Intel Inside as a stampof quality, but it gets worse by Tan.
Also saying that it's too latefor Intel to catch up with
Nvidia in AI training chips.
He's basically giving up on the mostprofitable part of the chip business.
(01:33):
The numbers tell the story.
Intel is worth $100 billion today,half of what it was 18 months ago,
but Nvidia hit $4 trillion thisweek, and the bleeding continues.
Intel lost $16 billion lastquarter despite spending more on
(01:55):
R&D than Nvidia and a MD combined.
So what went wrong?
One guess is that Inteljust got comfortable.
They made tiny chip improvements fordecades while dominating the industry.
They didn't see AMD comingwith better processors.
They missed the AI boom thatmade Nvidia Giant and their own
(02:17):
factories became a liability.
When competitors started using betterfoundries, they believed their own pr.
now Intel is laying off thousandsand outsourcing a third of their
production to the same Taiwanese companythat makes chips for their rivals.
Tan's plan, focus on the AI thatruns your devices instead of
(02:39):
chasing NVIDIA's cloud business.
It's smaller, but Intel mightactually have a shot there.
He's called this a marathon turnaround,where Intel needs to be humble.
For a company that used to dictate howthe chip industry worked, that's gonna
be quite a change and quite a challenge.
We've all heard the hype about datacentres and the economic boom that they
(03:01):
can bring to communities, but the realityfor people living next to these massive
facilities tells a very different story.
The BBC did a story on Beverly Morrisin Georgia who lives 400 yards from
a Meta data center, and she toldthe BBC, I can't drink the water.
I can't live in my home with halfof my home functioning and no water.
(03:23):
And her water woes happened justafter the new data center moved in.
Now the company denies it and pullsout studies that back up their
claims that they aren't responsible.
But yet, Maurice's claim is credible.
These facilities can guzzle millionsof gallons of water daily to cool their
servers rivaling entire town's usages.
(03:47):
But the problems go beyond water issues.
those counties that thought datacentres would create jobs and offered
them tax concessions and subsidies inreturn have been gravely disappointed.
The reality is that most facilitiesemploy only a handful of people.
Some estimates are between six and 30.
That's it.
(04:08):
Why they're highly
,But they consume massive amounts of not just water, but also electricity,
taxing local utilities and requiringexpensive upgrades to water and power
transmission and generation and still.
Often have to run diesel backupgenerators that pump out toxic emissions.
(04:29):
Recent studies found that airpollution from US data centres could
cause between 1.5 and as high aspotentially $6 billion in public health
damages in 2023 alone, some thinkit could hit $20 billion by 2030.
Bringing health issues fromfine particulates that will
trigger asthma and heart attacks.
(04:51):
Now back to the story of BeverlyMorris, there is the question of
what happens to property values.
While some areas see increases fromdevelopment residents near industrial
scale facilities could see their homevalues plummet, especially if stories
like Morris's are widely spread, andit's not just water and pollution.
There are reports of constanthumming, diesel generator
(05:14):
testing and industrial areas.
To be frank.
Don't exactly screamdesirable neighborhood.
And maybe that's why many centresare being built and depressed or
lower income communities whereproperty values are already lower and
residents have less political powerto fight back, and the problems may
(05:34):
get worse as these centres struggleto get enough power to keep running.
Elon Musk's new mega center hasbeen running backup generators
until they can wait for it, assemblean entirely new power plant that
they were importing from Europe.
Virginia just had to grant emergencypermits, allowing nearly 300
data centres to run their backupgenerators more frequently because
(05:57):
the power grid can't keep up.
And residents are feeling that.
They're basically becoming Guinea pigsfor an industry that seems to them to
be prioritizing digital infrastructureover their community health.
Companies promise they're investingin cleaner technology, but the
current pace of growth is outstrippingany environmental improvements.
(06:19):
But even though the data center boomhas potential issues, the demand for
processing means it's unlikely to slowdown regardless of environmental issues.
OpenAI's $3 billion deal to buy AI coding.
Startup Windsurf collapsed onFriday, but it also has a twist.
Google immediately stepped in and hiredWindsurf, CEO, and the top talent in what
(06:44):
they're calling a reverse acquisition.
Don't buy the company, take the top talentand license their products and services.
But the deal and Google'stactics reveal two major cracks
forming in the AI industry.
First, we've talked about therift between Microsoft and OpenAI,
but it appears to be growing.
(07:05):
The Windsurf deal was a major stickingpoint in their contract negotiations.
Microsoft currently has access toall of OpenAI's intellectual property
under their partnership agreement.
But OpenAI reportedly didn't wantMicrosoft getting Windsurf AI coding
technology, perhaps because theywant to compete with Microsoft.
(07:27):
Think about this for a second.
OpenAI is basically tryingto hide technology from its
biggest backer and partner.
This is not the behavior of companieswith a warm growing relationship, and
there have been reports that OpenAIexecutives have even discussed accusing
Microsoft of anti-competitive behaviorand filing federal antitrust complaints
(07:49):
for a company that owes its existenceto Microsoft's billions in funding.
That's pretty remarkable,but it's not an empty threat.
Antitrust is a big deal.
Google has been through at leasttwo court cases in the US that
threatened to split up the company,and there are more threats in Europe.
These tech giants are gettingnervous about antitrust scrutiny.
(08:10):
Remember, Google didn'tactually buy windsurf.
They just hired the key peopleand licensed the technology.
This is the new playbook.
Get the talent and the techwithout the full acquisition that
might trigger regulatory review.
It's not a new strategy for Google.
They did the same thing with character ai.
CEO Microsoft did the same withInflections, founder and these
(08:33):
reverse acquisitions let them buildAI dominance without technically
creating monopolies, at least on paper.
But here's what nobody's talking about.
Windsurf and its shareholders areprobably gonna get screwed by this deal.
Other startups that went throughthese talent raids, saw their
businesses collapse Scale.
AI lost customers after meta poached.
(08:55):
Their people inflection had to completelypivot after Microsoft gutted their team.
So the bottom line, the biggest AIcompanies are playing a shell game,
avoiding antitrust rules while buildingoligopoly control, and their partnerships
are falling apart as they partner, butcompete viciously for AI dominance.
(09:16):
Scientific American reported on a secretmeeting where 30 top mathematicians
tried to outsmart OpenAI's latestAI with the hardest math problems.
They could think up.
The AI blew them away, and we're talkingabout problems that would take human
experts weeks or months to solve AIknocked them out in a few minutes.
(09:38):
One mathematician reportedly gavethe AI an unsolved problem from
number theory, PhD level stuff.
He watched the AI spend two minutesreading up on the field, then say
it wanted to try an easier versionfirst to learn the approach.
Five minutes later, it's solvedthe original problem and basically
told him no citation needed.
(10:00):
I figured this out myself.
Apparently this AI reportedlyOpenAI's oh four Mini, not only
has ability, it also has attitude.
After two days of trying, these leadingmathematicians could only manage to create
10 questions The AI couldn't answer.
One researcher said working withit was like having a very, very,
(10:22):
very good graduate student.
Nah, actually better, and thiscan't be dismissed as fancy.
pattern matching.
The AI appeared to show evidence of realreasoning, breaking down complex problems,
learning from simpler ones applyingthat knowledge to harder challenges.
That doesn't mean we should stop lookingcritically at reports of what AIS can do.
(10:45):
All too often companies report thegood news or skew their results.
Elon Musk recently claimed thathis Grok 4 was smarter than PhDs.
And in terms of knowledge retrieval,it probably is, but we'd have to see
where it would stand on an evaluationlike this one, confidential, no
previous learning involved, andthe experts posing the challenges
(11:10):
where we are with artificial generalIntelligence is still an open
question, but we are either at thepoint or getting close to the point
where AI is simply smarter thanthe smartest humans in many areas.
the next question is, how far along arewe on what some of these mathematicians
(11:31):
claimed to see genuine reasoningand not just sophisticated copying?
And finally an update on thatRogers starlink satellite service
we talked about, somebody hasspotted it in the wilds of Canada.
A Reddit user in northwest Ontarioposted a screenshot showing
Roger's satellite on their phone.
(11:51):
They couldn't connect yet.
Probably not enabled on their account,but their phone could see the network.
So it appears that Rogers istesting their Starlink partnership.
The service is proposed to work withany LTE phone giving you coverage
where there's zero cell service, andit will start with SMS and emergency
calling, but then expand a full datacoverage anywhere Starlink satellite's
(12:13):
reach, which is basically everywhere.
T-Mobile is doing the same thingin the US for 10 to $20 a month.
And unlike Apple's emergencySOS, that only works on
iPhone 14 and newer Starlink.
We'll work with any LTE phone globally.
Those of us who live in rural areaswhere there are huge gaps in cellular
(12:33):
coverage, or those who like to exploreCanada's vast outdoor wilderness
will certainly welcome this service.
And that's our show.
a note to say thanks to thoseof you who came to see me at the
Minden Book of Palooza, where I waspleased to present my new novel,
Elisa to an enthusiastic audience.
(12:53):
And share the floor with notablewriters like Robert Sawyer, a Canadian
sci-fi legend, and if you miss theshow, but still wanna read about.
Ai, romance, mystery,and maybe even some hope.
Check it out@Elisabook.com.
That's E-L-I-S-A book.com,Elisa book.com, or you could
(13:14):
find me on Amazon as Elisa Jim.
Love Great reading for the Summeron the doc, along with your
favorite Robert Sawyer book.
I highly recommend his new book, theDownloaded, And in that spirit, whatever
you're reading this summer, and Sawyerfans will understand this one live long
and prosper and have a marvelous Monday.
Intel's, CEO delivers bad news.
Data centres are coming inconflict with their neighbors.
OpenAI's, $3 billion acquisition ofWindsurf Falls apart at the last minute.
And AI shocks some of the world's bestmathematicians in a confidential review.
Welcome to Hashtag Trending.
(00:24):
I'm your host, Jim Love.
Let's get into it.
Intel's, CEO made a stunning admissionin a speech to his employees.
Intel has fallen off the list of the top10 semiconductor companies, and it's too
late for Intel to catch up with Nvidia.
Anyone listening to or readingthe tech news would be aware that
(00:46):
Intel was no longer the leader, butjust how far they've fallen was a
shock . Lip-Bu Tan the CEO made thecomments during an internal meeting
this week and they got leaked.
He said 20, 30 years ago.
We were really the leader.
Now, I think the world has changed.
We're not in the top 10semiconductor companies.
(01:11):
Now that's stunning from a companythat used to own the chip industry.
Do you remember Intel Inside as a stampof quality, but it gets worse by Tan.
Also saying that it's too latefor Intel to catch up with
Nvidia in AI training chips.
He's basically giving up on the mostprofitable part of the chip business.
(01:33):
The numbers tell the story.
Intel is worth $100 billion today,half of what it was 18 months ago,
but Nvidia hit $4 trillion thisweek, and the bleeding continues.
Intel lost $16 billion lastquarter despite spending more on
(01:55):
R&D than Nvidia and a MD combined.
So what went wrong?
One guess is that Inteljust got comfortable.
They made tiny chip improvements fordecades while dominating the industry.
They didn't see AMD comingwith better processors.
They missed the AI boom thatmade Nvidia Giant and their own
(02:17):
factories became a liability.
When competitors started using betterfoundries, they believed their own pr.
now Intel is laying off thousandsand outsourcing a third of their
production to the same Taiwanese companythat makes chips for their rivals.
Tan's plan, focus on the AI thatruns your devices instead of
(02:39):
chasing NVIDIA's cloud business.
It's smaller, but Intel mightactually have a shot there.
He's called this a marathon turnaround,where Intel needs to be humble.
For a company that used to dictate howthe chip industry worked, that's gonna
be quite a change and quite a challenge.
We've all heard the hype about datacentres and the economic boom that they
(03:01):
can bring to communities, but the realityfor people living next to these massive
facilities tells a very different story.
The BBC did a story on Beverly Morrisin Georgia who lives 400 yards from
a Meta data center, and she toldthe BBC, I can't drink the water.
I can't live in my home with halfof my home functioning and no water.
(03:23):
And her water woes happened justafter the new data center moved in.
Now the company denies it and pullsout studies that back up their
claims that they aren't responsible.
But yet, Maurice's claim is credible.
These facilities can guzzle millionsof gallons of water daily to cool their
servers rivaling entire town's usages.
(03:47):
But the problems go beyond water issues.
those counties that thought datacentres would create jobs and offered
them tax concessions and subsidies inreturn have been gravely disappointed.
The reality is that most facilitiesemploy only a handful of people.
Some estimates are between six and 30.
That's it.
(04:08):
Why they're highly
,But they consume massive amounts of not just water, but also electricity,
taxing local utilities and requiringexpensive upgrades to water and power
transmission and generation and still.
Often have to run diesel backupgenerators that pump out toxic emissions.
(04:29):
Recent studies found that airpollution from US data centres could
cause between 1.5 and as high aspotentially $6 billion in public health
damages in 2023 alone, some thinkit could hit $20 billion by 2030.
Bringing health issues fromfine particulates that will
trigger asthma and heart attacks.
(04:51):
Now back to the story of BeverlyMorris, there is the question of
what happens to property values.
While some areas see increases fromdevelopment residents near industrial
scale facilities could see their homevalues plummet, especially if stories
like Morris's are widely spread, andit's not just water and pollution.
There are reports of constanthumming, diesel generator
(05:14):
testing and industrial areas.
To be frank.
Don't exactly screamdesirable neighborhood.
And maybe that's why many centresare being built and depressed or
lower income communities whereproperty values are already lower and
residents have less political powerto fight back, and the problems may
(05:34):
get worse as these centres struggleto get enough power to keep running.
Elon Musk's new mega center hasbeen running backup generators
until they can wait for it, assemblean entirely new power plant that
they were importing from Europe.
Virginia just had to grant emergencypermits, allowing nearly 300
data centres to run their backupgenerators more frequently because
(05:57):
the power grid can't keep up.
And residents are feeling that.
They're basically becoming Guinea pigsfor an industry that seems to them to
be prioritizing digital infrastructureover their community health.
Companies promise they're investingin cleaner technology, but the
current pace of growth is outstrippingany environmental improvements.
(06:19):
But even though the data center boomhas potential issues, the demand for
processing means it's unlikely to slowdown regardless of environmental issues.
OpenAI's $3 billion deal to buy AI coding.
Startup Windsurf collapsed onFriday, but it also has a twist.
Google immediately stepped in and hiredWindsurf, CEO, and the top talent in what
(06:44):
they're calling a reverse acquisition.
Don't buy the company, take the top talentand license their products and services.
But the deal and Google'stactics reveal two major cracks
forming in the AI industry.
First, we've talked about therift between Microsoft and OpenAI,
but it appears to be growing.
(07:05):
The Windsurf deal was a major stickingpoint in their contract negotiations.
Microsoft currently has access toall of OpenAI's intellectual property
under their partnership agreement.
But OpenAI reportedly didn't wantMicrosoft getting Windsurf AI coding
technology, perhaps because theywant to compete with Microsoft.
(07:27):
Think about this for a second.
OpenAI is basically tryingto hide technology from its
biggest backer and partner.
This is not the behavior of companieswith a warm growing relationship, and
there have been reports that OpenAIexecutives have even discussed accusing
Microsoft of anti-competitive behaviorand filing federal antitrust complaints
(07:49):
for a company that owes its existenceto Microsoft's billions in funding.
That's pretty remarkable,but it's not an empty threat.
Antitrust is a big deal.
Google has been through at leasttwo court cases in the US that
threatened to split up the company,and there are more threats in Europe.
These tech giants are gettingnervous about antitrust scrutiny.
(08:10):
Remember, Google didn'tactually buy windsurf.
They just hired the key peopleand licensed the technology.
This is the new playbook.
Get the talent and the techwithout the full acquisition that
might trigger regulatory review.
It's not a new strategy for Google.
They did the same thing with character ai.
CEO Microsoft did the same withInflections, founder and these
(08:33):
reverse acquisitions let them buildAI dominance without technically
creating monopolies, at least on paper.
But here's what nobody's talking about.
Windsurf and its shareholders areprobably gonna get screwed by this deal.
Other startups that went throughthese talent raids, saw their
businesses collapse Scale.
AI lost customers after meta poached.
(08:55):
Their people inflection had to completelypivot after Microsoft gutted their team.
So the bottom line, the biggest AIcompanies are playing a shell game,
avoiding antitrust rules while buildingoligopoly control, and their partnerships
are falling apart as they partner, butcompete viciously for AI dominance.
(09:16):
Scientific American reported on a secretmeeting where 30 top mathematicians
tried to outsmart OpenAI's latestAI with the hardest math problems.
They could think up.
The AI blew them away, and we're talkingabout problems that would take human
experts weeks or months to solve AIknocked them out in a few minutes.
(09:38):
One mathematician reportedly gavethe AI an unsolved problem from
number theory, PhD level stuff.
He watched the AI spend two minutesreading up on the field, then say
it wanted to try an easier versionfirst to learn the approach.
Five minutes later, it's solvedthe original problem and basically
told him no citation needed.
(10:00):
I figured this out myself.
Apparently this AI reportedlyOpenAI's oh four Mini, not only
has ability, it also has attitude.
After two days of trying, these leadingmathematicians could only manage to create
10 questions The AI couldn't answer.
One researcher said working withit was like having a very, very,
(10:22):
very good graduate student.
Nah, actually better, and thiscan't be dismissed as fancy.
pattern matching.
The AI appeared to show evidence of realreasoning, breaking down complex problems,
learning from simpler ones applyingthat knowledge to harder challenges.
That doesn't mean we should stop lookingcritically at reports of what AIS can do.
(10:45):
All too often companies report thegood news or skew their results.
Elon Musk recently claimed thathis Grok 4 was smarter than PhDs.
And in terms of knowledge retrieval,it probably is, but we'd have to see
where it would stand on an evaluationlike this one, confidential, no
previous learning involved, andthe experts posing the challenges
(11:10):
where we are with artificial generalIntelligence is still an open
question, but we are either at thepoint or getting close to the point
where AI is simply smarter thanthe smartest humans in many areas.
the next question is, how far along arewe on what some of these mathematicians
(11:31):
claimed to see genuine reasoningand not just sophisticated copying?
And finally an update on thatRogers starlink satellite service
we talked about, somebody hasspotted it in the wilds of Canada.
A Reddit user in northwest Ontarioposted a screenshot showing
Roger's satellite on their phone.
(11:51):
They couldn't connect yet.
Probably not enabled on their account,but their phone could see the network.
So it appears that Rogers istesting their Starlink partnership.
The service is proposed to work withany LTE phone giving you coverage
where there's zero cell service, andit will start with SMS and emergency
calling, but then expand a full datacoverage anywhere Starlink satellite's
(12:13):
reach, which is basically everywhere.
T-Mobile is doing the same thingin the US for 10 to $20 a month.
And unlike Apple's emergencySOS, that only works on
iPhone 14 and newer Starlink.
We'll work with any LTE phone globally.
Those of us who live in rural areaswhere there are huge gaps in cellular
(12:33):
coverage, or those who like to exploreCanada's vast outdoor wilderness
will certainly welcome this service.
And that's our show.
a note to say thanks to thoseof you who came to see me at the
Minden Book of Palooza, where I waspleased to present my new novel,
Elisa to an enthusiastic audience.
(12:53):
And share the floor with notablewriters like Robert Sawyer, a Canadian
sci-fi legend, and if you miss theshow, but still wanna read about.
Ai, romance, mystery,and maybe even some hope.
Check it out@Elisabook.com.
That's E-L-I-S-A book.com,Elisa book.com, or you could
(13:14):
find me on Amazon as Elisa Jim.
Love Great reading for the Summeron the doc, along with your
favorite Robert Sawyer book.
I highly recommend his new book, theDownloaded, And in that spirit, whatever
you're reading this summer, and Sawyerfans will understand this one live long
and prosper and have a marvelous Monday.
Popular Podcasts
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.